For analysis of our illumina and duet samples using Griffin I have used input files from Griffin paper analyses, and the ITSFASTR pipeline. We need to clone those repos:
git clone git@github.com:mouliere-lab/ITSFASTR.git
git clone git@github.com:adoebley/Griffin_analyses.gitHere we will use data previously used to assess fragmentomics comparability between duet and illumina WGS (DS-417).
bash bin/get_data.shYou will need to have the reference genome in place to run the pipeline, in the current
config files I have specified its location as /scratch/reference/GRCh38Decoy.fa, it must
also be fai indexed using samtools.
Create sample sheets:
bash bin/create_samples.sh data/duet duet_samples.yaml
mv duet_samples.yaml snakemakes/griffin_GC_and_mappability_correction/config/
bash bin/create_samples.sh data/illumina ilm_samples.yaml
mv ilm_samples.yaml snakemakes/griffin_GC_and_mappability_correction/config/conda env create -n griffin -f envs/griffin.yaml
conda activate griffinDifferent config files are required to run Griffin in normal mode or long-read mode, which requires using the modified profiling script from ITSFASTR. Long read mode is used for duet in order to use the read length as the fragment length.
for seq in "duet" "ilm"; do
mkdir -p "results_${seq}"
cd snakemakes/griffin_GC_and_mappability_correction/
snakemake \
-s griffin_GC_and_mappability_correction.snakefile \
-j 2 \
--configfile "../../config_${seq}.yaml" \
--configfile "../../${seq}_samples.yaml"
cp "results/samples.GC.yaml" ../../snakemake/griffin_nucelosome_profiling/
mv results/* "../../results_${seq}"
cd ../griffin_nucleosome_profiling
snakemake \
-s griffin_nucelosome_profiling.snakefile -j 2 \
--configfile "../../config_${seq}.yaml" \
--configfile "../../${seq}_samples.yaml"
mv results/* "../../results_${seq}"## Analysis
Analysis of outputs was performed in CA-802.